Skip to content

Latest commit

 

History

History
12 lines (8 loc) · 464 Bytes

README.md

File metadata and controls

12 lines (8 loc) · 464 Bytes

torch-mnist-batch-normalization

Torch7 batch normalization example on MNIST dataset

I use torch7 to test the techniq from the Batch Normalization paper, section 4.1.

The main difference is the data, MNIST dataset I use has pictures of size 32x32px.

Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift by Sergey Ioffe, Christian Szegedy (11 Feb 2015) http://arxiv.org/abs/1502.03167

Example results